Planet Neo4j
Ever since Facebook promoted its “graph search” methodology, lots of people in our industry have been waking up to the fact that graphs are über-cool. Thanks to the powerful query possibilities, people like Facebook, Twitter, LinkedIn, and let us not forget, Google have been providing us with some of the most amazing technologies. Specifically, the power of the “social network” is tempting
The current version of Neo4j has two different types of index: named indexes and automatic indexes. The Cypher query examples throughout the Graph Databases book use named indexes. In this query, for example, venue, city and author are named node indexes:
START theater=node:venue(name='Theatre Royal'),
newcastle=node:city(name='Newcastle'),
bard=node:author(lastname='Shakespeare')
MATCH (newcastle)<-[:STREET|CITY*1..2]-(theater)
<-[:VENUE]-()-[:PERFORMANCE_OF]->()-[:PRODUCTION_OF]->
(play)<-[:WROTE_PLAY]-(bard)
RETURN DISTINCT play.title AS playThere’s a problem, however: data created using a Cypher CREATE statement won’t be indexed in a named index. This has led to some confusion for anyone wanting to code along with the examples: if you use the CREATE statements as published, the query examples won’t work.
(You can find out more about Neo4j indexes in the documentation.)
If you want to code along with the examples, the easiest thing to do is create the data using Cypher and automatic indexes. You’ll need to modify the queries to use the automatic indexes, but the changes aren’t complicated.
Configure Automatic Indexes
The examples below show how to run the Shakespeare examples (pp.47-50) using automatic indexes.
Before you create any data, you’ll need to configure the automatic indexes. The approach you take will depend on whether you are running Neo4j in embedded or server mode.
Embedded mode
If running in embedded mode, configure the database at startup:
GraphDatabaseService graphDb = new GraphDatabaseFactory().
newEmbeddedDatabaseBuilder( storeDirectory ).
setConfig( GraphDatabaseSettings.node_keys_indexable, "name,lastname" ).
setConfig( GraphDatabaseSettings.node_auto_indexing, "true" ).
newGraphDatabase();Server mode
If running in server mode and executing Cypher statements through the console or using the REST API, add the following to the conf/neo4j.properties file.
node_auto_indexing=true
node_keys_indexable=name,lastnameModify the Queries
The Shakespeare queries should then be modified as follows. In each case we’ve replaced a named node index (e.g. node:author) with an automatic node index (node:node_auto_index).
p.47:
START theater=node:node_auto_index(name='Theatre Royal'),
newcastle=node:node_auto_index(name='Newcastle'),
bard=node:node_auto_index(lastname='Shakespeare')
MATCH (newcastle)<-[:STREET|CITY*1..2]-(theater)
<-[:VENUE]-()-[:PERFORMANCE_OF]->()-[:PRODUCTION_OF]->
(play)<-[:WROTE_PLAY]-(bard)
RETURN DISTINCT play.title AS playp.48:
START theater=node:node_auto_index(name='Theatre Royal'),
newcastle=node:node_auto_index(name='Newcastle'),
bard=node:node_auto_index(lastname='Shakespeare')
MATCH (newcastle)<-[:STREET|CITY*1..2]-(theater)
<-[:VENUE]-()-[:PERFORMANCE_OF]->()-[:PRODUCTION_OF]->
(play)<-[w:WROTE_PLAY]-(bard)
WHERE w.year > 1608
RETURN DISTINCT play.title AS playp.49:
START theater=node:node_auto_index(name='Theatre Royal'),
newcastle=node:node_auto_index(name='Newcastle'),
bard=node:node_auto_index(lastname='Shakespeare')
MATCH (newcastle)<-[:STREET|CITY*1..2]-(theater)
<-[:VENUE]-()-[p:PERFORMANCE_OF]->()-[:PRODUCTION_OF]->
(play)<-[:WROTE_PLAY]-(bard)
RETURN play.title AS play, count(p) AS performance_count
ORDER BY performance_count DESCp.50:
START bard=node:node_auto_index(lastname='Shakespeare')
MATCH (bard)-[w:WROTE_PLAY]->(play)
WITH play
ORDER BY w.year DESC
RETURN collect(play.title) AS plays
